Abstract A growing body of literature demonstrates dramatic disparities in health between lesbian, gay and bisexual (LGB) and heterosexual youth, including problem substance use, HIV risk behaviors and suicide. In 2011 the Institute of Medicine called for additional research with this population, prioritizing health disparities research as essential for building a solid evidence base in LGB heath. The IOM report highlights the importance of multilevel social influences on the lives of LGB people, specifying the role of social structures such as families, schools, workplaces, religious institutions and community organizations as critical subjects of study to expand our understanding of the determinants of health in this population. The proposed R01 led by a new investigator in response to PA-12-111, directly addresses this need. The goal of this study is to broaden and deepen our understanding of the family, peer, school and community environments that protect young LGB people from involvement in high risk health behaviors, including substance use, HIV risk behaviors and suicide behaviors. This research aims to 1) develop a theoretically grounded approach to promoting health among LGB adolescents based on in-depth knowledge of their community and school environments, and 2) link environmental data, collected using the Inventory, with existing population-based student data to identify factors at the individual, family, peer, school and community levels that protect LGB youth from involvement in health risk behaviors. Go-along interviews with 72 youth in diverse locations will be used to elicit in-depth information on LGB adolescents' perceptions of supportive elements in their schools and communities; this information, in conjunction with published literature, expert review, and psychometric testing, will be used to create an LGB Environment Inventory to characterize policies, programs, resources and other supports for LGB youth that exist in these settings. The Inventory will then be used to measure indicators of support in 120 communities in Minnesota, British Columbia and Massachusetts, using publicly available materials (e.g. websites) and brief contacts with key informants (Aim 1). These community-level data will then be linked with existing student survey data from approximately 3,600 LGB adolescents in these same communities, which will include information about family, peer and individual supports, as well as health behaviors and demographic information (Aim 2). The following hypotheses will be tested: a) higher LGB environment scores for the community and school will be protective against health risk behaviors among LGB youth; b) greater family connectedness and support and a more supportive peer environment will be protective against health risk behaviors among LGB youth, and c) family, peer and individual-level factors will moderate the associations between the LGB environment (community and school) and the health risk behaviors of interest.